For example, automobile internal combustion engines include a catalyst installed in an exhaust system to purify exhaust gas. Some such catalysts have an O2 storage capability. When the air-fuel ratio of exhaust gas flowing into the catalyst is higher than a theoretical air-fuel ratio (stoichiometry), that is, the air-fuel ratio is indicative of a lean state, a catalyst with the O2 storage capability absorbs excess oxygen present in the exhaust gas. When the air-fuel ratio of the exhaust gas is lower than the stoichiometry, that is, the air-fuel ratio is indicative of a rich state, the catalyst releases the absorbed oxygen. For example, a gasoline engine controls the air-fuel ratio so as to set the air-fuel ratio of the exhaust gas flowing into the catalyst close to the stoichiometry. When the gasoline engine uses a three-way catalyst with the O2 storage capability, even if the actual air-fuel ratio deviates slightly from the stoichiometry, the three-way catalyst can absorb such a deviation of the air-fuel ratio by the oxygen absorbing and releasing effect.
On the other hand, a degraded catalyst has a reduced conversion efficiency. The degree of degradation of the catalyst correlates with the degree of a decrease in the O2 storage capability. Thus, detection of a reduced O2 storage capability enables degradation or abnormality of the catalyst to be detected. In general, the following method is adopted far the detection: active air-fuel ratio control is performed to alternately control the air-fuel ratio in an area located upstream of the catalyst between a rich side and a lean side, the amount of oxygen absorbed and released by the catalyst is measured during the lean control and the rich control, and the abnormality of the catalyst is diagnosed based on the amount of oxygen (see, for example, Patent Literature 1).
According to this Cmax method, a post-catalyst sensor is provided which detects the exhaust air-fuel ratio in an area located downstream of the catalyst. When an output from the past-catalyst sensor is inverted, then simultaneously the lean control and the rich control are switched from one to another or vice versa, and measurement of the amount of oxygen is ended.
However, when the amount of oxygen is measured, the amount of oxygen that is actually not absorbed or released is also measured, disadvantageously resulting in measurement errors. In particular, according to the conventional Cmax method, an error rate measured immediately before the inversion of the output from the post-catalyst sensor is higher when the catalyst is abnormal than when the catalyst is normal. This enhances the tendency to make the measured value larger than the real value. Then, a catalyst that is actually abnormal may be erroneously diagnosed to be normal. The Cmax method may also preclude an increase in a difference in the measured value of the amount of oxygen between the normal state and abnormal state of the catalyst. In particular, if this difference is originally small, the catalyst may not sufficiently accurately be diagnosed.
Thus, the present invention has been developed in view of the above-described circumstances. An object of the present invention is to provide a catalyst abnormality diagnosis apparatus that can improve the diagnosis accuracy to suppress erroneous diagnoses.